Decadal drought deaccelerated the increasing trend of annual net primary production in tropical or subtropical forests in southern China
R A
P
WantongWang,, JinxiaWang, Xingzhao Liu, Guoyi Zhou & JunhuaYan
(NPP) of global forests have varied both spatially and temporally, and that warming has increased the NPP for many forests. However, other factors, such as available soil water for plant growth, could limit of tropical or subtropical forests in southern China to warming and drought stress over the past three
a (standardized) during the study period, decade. This deacceleration was primarily caused by a decrease in available soil water which resulted from warming (mainly occurring in winter and autumn) and the changes in rainfall pattern. The result indicates that intensifying drought stress would limit future increases of forest NPP in southern China.
Net primary production (NPP) is an important measure of terrestrial ecosystem functioning1. NPP represents the ability of terrestrial plants to x atmospheric CO2, and it is also a key ux of the global carbon cycle2. Climate warming directly inuences terrestrial NPP by changing the carbon uptake capacity of leaves; it also indirectly aects NPP by changing soil water availability3 and by prolonging the growing season4.
Temperature, water and radiation interact to impose complex and variable limitations on NPP in dierent parts of the world5. Many studies have revealed temporal changes in terrestrial ecosystem NPP in response to continental or global scale climate change59. In the temperature-limited regions at high latitudes, such as North America10 and Northern China11,12, warming extended the plant growing season by promoting earlier plant growth and their NPP. Nevertheless, in predominantly radiation-limited regions over the middle and low latitudes, such as the Amazon basin1315, an increase in solar radiation, owing to declining cloud cover, is the most likely explanation for increased NPP of tropical forests.
Chinas tropical and subtropical evergreen broadleaved forests (TEBF) are located in the region between 20N to 31N and 101E to 122E, an area which covers more than 26% of Chinas land surface (Fig.1). Over the past several decades, dramatic climatic changes have occurred in this region. Mean annual temperature (MAT) has increased at a rate of 0.2C decade1 from 1952 to 2011 (this value is higher than the 0.17C decade1 on a global scale recorded between 1948 and 2010)6, whilst mean annual precipitation (MAP) has remained unchanged16. Furthermore, rainfall patterns have shied towards more severe storms. As a result, the soils have become drier
SCIENTIFIC REPORTS
1
www.nature.com/scientificreports/
Figure 1. Locations of the ve eld stations in Chinas tropical and subtropical evergreen broadleaved forests (Map created using ARCGIS 10.1 soware by rst author, URL: http://www.esri.com). The ve eld stations are TT-Tiantongshan, DH-Dinghushan, HT-Huitong, BN-Xishuangbanna, and AL-Ailaoshan.
Site
name Forest type Dominant species
Forest
age(year) Location Altitude(m) MAT(C) MAP(mm)
AL
Subtropical
evergreen
broadleaved
forest
Lithocarpus xylocarpus,
Castanopsis rufescens
>250
10101E, 2432N
2488
11.3
1982
BN
Tropical
seasonal rain
forest
Pometia pinnata,
Terminalia myriocarpa,
Gironniera subaequalis
>100
101121E, 2157 40N
750
21.5
1557
DH
Monsoon
evergreen
broadleaved
forest
Castanopsis chinensis,
Schima superba,
Michelia odora
>400
1123222E, 231011N
220
21.0
1956
HT
Subtropical
evergreen
broadleaved
forest
Castanopsis hystrix,
Cyclobalanopsis
>250
10936E, 2650N
350
16.5
1300
TT
Subtropical
evergreen
broadleaved
forest
Castanopsis fargesii
>200
1214712E, 294829N
196
17.1
1408
Table 1. Description of the ve sites.
and extreme hydrological events (e.g. droughts and oods) have become more frequent17. Overall, such rapid warming and change in rainfall patterns has led to an increase in drought stress, thereby reducing tree diameter growth18, enhancing tree mortality19,20, and decreasing the biomass carbon sink21. Zhou et al., investigating TEBF biome changes based on the observations from established permanent plots over the past four decades, reported that TEBF ecosystems have undergone a transition from cohorts with few large individuals to ones with a larger number of smaller individuals. They suggested that these subtropical forests were under threat due to their lack of resilience to these changes22. The reorganization of these biomes was the result of regional warming and associated soil drying16, ndings which were consistent with results from drought experiments in the Amazon forest23.
However, few studies have examined the impact of climate change on TEBF in southern China. There are still some urgent questions which need to be answered, such as (1) how will the TEBF ecosystems NPP changes under such a long-term warming and the consequences of drought stress; and (2) what are the main limiting climate factors aecting NPP in this region, and what impact do they have? In this study, we compiled and analyzed NPP data based on the AVHRR GloPEM NPP data24 (19812000) and the MOD17A3 NPP data25 (20012012), and analyzed the corresponding climatic variables and soil moisture across ve forest eld research stations in the TEBF (Fig.1 and Table1) from 1981 to 2012. Pearson correlation analyses were performed to determine the climatic factors that caused signicant changes in the NPP. Finally, we synthesized the latest ndings on the eect of drought stress on TEBF ecosystems and carbon cycling.
Results
Annual or decadal changes in climate variables. Table2 shows the change in climate variables for the study region from 1981 to 2012. Overall, there was a signicant increase at the rate of 0.057C a1 (standardized) in MAT and of 0.067 d a1 (standardized) in annual days without rainfall. For mean annual relative humidity (MARH), annual days with small rainfall, and soil water content in the top 50cm, there was a signicant decrease at the rate of 0.062% a1 (standardized), 0.048 d a1 (standardized), and 0.063 mm a1 (standardized), respectively. For MAP, no signicant change was identied.
However, changes observed in dierent decades were dierent from annual changes for the climate variables and soil water content over the study period (Fig.2). Firstly, in the dierent decades, the changes for the climate variables and soil water content were dierent. Results for MAP, for example, showed a decrease at the rate of 0.078 mm a1 (standardized) in the rst decade (from 1981 to 1990), an increase at the rate of 0.017 mm a1 (standardized) in the second decade (from 1990 to 2000), and a decrease at the rate of 0.055 mm
SCIENTIFIC REPORTS
2
www.nature.com/scientificreports/
b SE R2 P n
MAP~ 0.001 0.008 <0.001 0.884 160 MAT 0.057 0.007 0.286 <0.001 160 MARH 0.062 0.007 0.314 <0.001 160 Top 50 soil water content 0.063 1.153 0.460 <0.001 49 Annual days without rainfall 0.067 0.007 0.373 <0.001 160 Annual days with small rainfall 0.048 0.008 0.193 <0.001 160 NPP 0.076 0.522 0.511 <0.001 160
Table 2. Trends in the standardized data representing annual climate variability and net primary production in tropical or subtropical evergreen broadleaved forests from 1981 to 2012. Note: signicant uptrend; signicant downtrend;~no signicant directional changes.
a1 (standardized) in the third decade (from 2001 to 2012). Compared with this, as already stated, there were no signicant trends for MAP over the study period (Table2). For MARH, and annual days with small rainfall, apart from signicant downtrends being identied from 1981 to 2012, similar characteristic changes to those identied for MAP in the three decades were identied. The signicance of climate variable change was also identied to be dierent. For example, a signicant uptrend was identied for MAT in the rst two decades, but in the third decade no signicant change was identied. For annual days without rainfall, a signicant uptrend was shown in the rst decade and the third decade, yet no signicant change was identied in the second decade. Overall, signicant changes for the climate variables mainly occurred in the rst and third decades, and a signicant downtrend for soil water content mainly occurred in the recent two decades.
Monthly or seasonal variations in trends of climate variables. Climate variables showed large variations within months or seasons (Fig.3). Overall, MAT increased the fastest in winter, followed by autumn, summer, and then spring, which led to a reduction in inter-seasonal dierences in temperature. At the monthly scale, the fastest increase in MAT occurred in February and October. MAP increased in the spring and summer and decreased in the autumn and winter, but these dierences were not signicant. This led to an increase in the uneven distribution of seasonal precipitation: the wet-season became wetter and the dry-season became drier. Annual days with small rainfall and annual days without rainfall showed opposite trends in which the former decreased more quickly in the autumn (mainly occurring in November) and winter (mainly occurring in January and February) than in spring and summer, while the latter increased faster in the same season and months than the former. This resulted in extended dry periods and severe droughts. MARH rates exhibited smaller dierences between seasons but had similar characteristics as displayed by annual days with small rainfall. Soil water content decreased faster in the spring and summer than in the winter and autumn, but the intensity of the decrease was contrary to that in MARH. This implied that the drought in the soil was delayed by two seasons compared to the atmosphere.
Annual NPP trends. As shown in Table2 and Fig.2, annual NPP increased at a rate of 0.076 g C m2 a2 (standardized) from 1981 to 2012, but with variations of the rate of increase between the three decades; 0.149g C m2 a1 in the 1980s, 0.106g C m2 a1 in the 1990s and 0.087g C m2 a1 for 20002012 (all values were standardized). Therefore, the increase of NPP was deaccelerated by 20.8% per decade over the study period.
Relationships between NPP trend and climate variables. Over the study period (1981 to 2012) there were strong correlations between annual NPP and climate variables, except for MAP (Table3). The results showed that annual NPP was strongly positively correlated with annual days without rainfall (r = 0.510, P < 0.001) and MAT (r = 0.508, P < 0.001); whereas it was signicantly negatively correlated with MARH (r = 0.491, P < 0.001), annual days with small rainfall (r = 0.352, P < 0.001), and the soil water content of the top 50cm (r=0.320, P=0.025).
For the seasonal patterns, the correlation between annual NPP and annual days without rainfall was strongly positive in autumn (r=0.330,P< 0.001) and winter (r=0.266,P< 0.001), while for annual days with small rainfall there was a negative correction (r=0.285, P<0.001; r=0.232, P= 0.004, respectively). Annual NPP was strongly positively correlated with MAT in the autumn (r=0.423, P< 0.001), winter (r=0.350, P<0.001), summer (r=0.324, P< 0.001), and spring (r=0.242, P= 0.002), and was strongly negatively correlated with MARH in the same four seasons (all P < 0.001). Overall, NPP was inuenced in the winter and autumn by changes in rainfall patterns (annual days without rainfall and annual days with small rainfall), while MAT and MARH had an equal inuence all year round.
As shown in Table4, annual NPP and climate variables (annual days without rainfall, MAT and MARH) were signicantly correlated, as determined by stepwise multi-regression analysis. This result was consistent with the above ndings. It was possible to explain the annual NPP variance that accounted for 39.9%. Standardized coefficients represented the contribution of the independent variables when explaining the dependent variable change. In this study, they represented the inuence of each climate variable on annual NPP. Therefore, annual days without rainfall were the most important factor for annual NPP, followed by MAT and then MARH. Within the total eects, 33.5% was due to MAT and 46% was due to changes in rainfall patterns (annual days without rainfall).
SCIENTIFIC REPORTS
3
www.nature.com/scientificreports/
Figure 2. Decadal trends of the standardized data representing climate variability and net primary production in the tropical or subtropical evergreen broadleaved forests over the past three decades. Soil water content (mm) was obtained from plots located at four stations (AL, BN, DH, HT).
SCIENTIFIC REPORTS
4
www.nature.com/scientificreports/
Figure 3. Variations in change rate of standardized data representing climate variables or net primary production at the monthly or seasonal scales in tropical or subtropical evergreen broadleaved forests from 1981 to 2012. Soil water content (mm) was obtained from the plots located at four stations (AL, BN, DH, HT). The slope obtained from a least square linear regression was used to estimate the rates (the letter a above histogram bar signies the coefficient by statistical signicance test (P<0.05).
SCIENTIFIC REPORTS
5
www.nature.com/scientificreports/
r
WIN SPR SUM AUT Annual
MAP 0.031 0.090 0.052 0.232 0.120 160 MAT 0.350** 0.242* 0.324** 0.423** 0.508** 160
MARH 0.474** 0.281** 0.435** 0.516** 0.491** 160 Top 50 soil water
content 0.193 0.237 0.321* 0.220 0.320* 49 Annual days without
rainfall 0.266** 0.110 0.121 0.330** 0.510** 160 Annual days with
small rainfall 0.232* 0.172* 0.064 0.285** 0.352** 160
Table 3. Pearsons correlation coefficients between the standardized data representing climate variability and annual net primary production in tropical or subtropical evergreen broadleaved forests from 1981to 2012. *P<0.05, **P<0.001.
Standardized
Coefficients P R2 SE n
Climate variables
Regression equation
y=0.354x1+0.256x2
0.157x3 0.031 0.012 0.399 0.765 160 x1 0.355 0.001x2 0.258 0.003x3 0.158 0.012x4 0.063 0.364x5 0.081 0.591x6 0.022 0.818
Table 4. Multiple-regression models of the standardized data representing annual net primary production for the tropical or subtropical evergreen broadleaved forests in China from 1981 to 2012. Dependent variables: y: NPP, Independent variables: x1: annual days without rainfall; x2: .MAT; x3: MARH; x4: MAP; x5: soil water content; x6: annual days with small rainfall.
Discussion
TEBFs are located in Chinas humid region where adequate water conditions ensure the sustained growth of forest vegetation. Our results showed an increasing trend for NPP at the rate of 0.076g C m2 a1 from 19812012 (about 7.9% from 1982 to 1999), which was greater than the global rate (about 6% from 1982 to 1999)8. However, within each of the three decades, the increasing rate of NPP showed a decreasing trend of 20.8% per decade. The dynamic change in NPP implied that forest ecosystems in this region were negatively impacted by climate change. Recent reports have conrmed this eect, and suggested that TEBF ecosystems experienced some signicant changes, such as reorganization of biomes16, reduction of tree diameter growth18, and enhanced tree mortality20,22, all of which were due to drought stress. Severe drought conditions can not only have a negative impact on tree growth in a moist tropical forest, it can also lead to a large volume of carbon emissions due to forest res and tree mortality23. On the other hand, our ndings show that the threats caused by climate change have yet to overcome the resilience of TEBF ecosystems22, as shown by the continued increase of NPP over the long-term period.
During 19812012, in the study region, the anisomerous change in climate variables between seasonal and decadal timescales has resulted in stronger droughts, especially soil droughts in the growing season (spring and summer), but this drought trend has mainly occurred in the most recent decade due to a lack of water (as shown in Fig.2, MAP, MARH, and soil water content signicantly decreasing between 20002012). Recently, NASA data has shown that there was a declining trend in plant growth worldwide due to droughts which has caused global plant productivity to only increase by 1% from 2000 to 20097.
Climate, at both regional and global scales, has a signicant eect on ecosystem processes26. Temperature, water and other factors interact to impose complex and varying limitations on NPP in dierent parts of the world27. When there is sufficient water, increases in temperatures and solar radiation can promote plant photo-synthesis to increase NPP, whereas drought stress can reduce photosynthesis and enhance autotrophic respiration, which together reduce NPP. Among all the climate variables we studied, annual days without rainfall and annual days with small rainfall control the precipitation pattern; their changes resulted in more drought events by gathering precipitation (more no-rainfall days) and increasing evapotranspiration (more sun radiation) to enhance the loss of water28. On the other hand, more radiation can promote plant photosynthesis to increase NPP when soil water is not signicantly limiting. For MAT, it is conrmed that a warming climate can lengthen the growing season4,29; furthermore, under drought stress, warming indirectly dampens photosynthesis as the plant tries to reduce water loss by reducing stomata conductance that causes a decline in NPP. Therefore, the interaction between temperature and precipitation conditions and the related changes in the climate caused the observed variations of NPP at our study sites.
n
SCIENTIFIC REPORTS
6
www.nature.com/scientificreports/
At high latitude or in semi-arid regions, rainfall is rare and unchanged, and temperature exerts the dominant control on plant growth by lengthening the growing seasons29; however, in low latitudes or humid regions, water availability has the dominant control on plant growth7. Our results also showed that annual days without rainfall and MAT exerted the most dominant inuences on NPP based on the correlation and multi-regression analysis in the study region. During the 1980s when soil water was not limiting, MARH and soil water content had less variation and warm winters enhanced plant growth in this region, and then during the 1990s, under weak drought stress, soil water content decreased. This change resulted in a reduction of the vegetations photosynthesis and the increasing rate of annual NPP declined by 28.9% than in the 1980s, although there being an increasing trend due to MAT. From 2000 to 2012, severe drought stress (MARH and soil water content decreased quickly) resulted in a continuous reduction in the increasing rate of NPP. Overall, the increasing rate of NPP was deaccelerated due to drought stress that was caused by anisomerous changes in seasonal warming and rainfall patterns in the region.
Overall, the anisomerous change in the rate of climate variables between months and seasons has resulted in warmer winters and stronger droughts in this region, especially during the growth season (spring and summer). It can also be predicted that as warming increases and rainfall patterns change in the future, stress due to drought conditions will continue to intensify and limit any future increase in forest NPP. This will therefore destabilize the forest carbon balance and weaken the terrestrial carbon sink in this region.
Materials and Methods
Site description. As the regional climax vegetation, Chinas TEBF dominates the directions for natural succession and articial restoration of all degraded forest ecosystems in the region30. A subtropical monsoon climate prevails in the study region due to its proximity to the South China Sea, the general air circulation, and the existence of the Qin-Tibet plateau. This climate has an MAP of 13002000mm; 80% of precipitation occurs in the wet season (April to September) and 20% occurs in the dry season (October to March). MAT and MARH are 15.0~21.5C and 78%, respectively30.
Across this region, since the 1970s, the Chinese Academy of Sciences and the Ministry of Science and Technology of the Peoples Republic of China have deployed a series of eld research stations. We selected ve of these eld research stations (Tiantongshan TT, Dinghushan DH, Huitong HT, Xishuangbanna BN, and Ailaoshan AL) (Fig.1 and Table1) to investigate the inuence of climate change on TEBF ecosystems NPP. It was conrmed that, up until the study date, the chosen ecosystems had not experienced articial management or natural catastrophic events (landslides, typhoons, forest res, etc.), and that no traces of such events can be found. In addition, for the whole plot, the soil proles are well preserved. Thus, the ve stations selected provided sufficient information about the response of the TEBF ecosystems to climate change under natural conditions.
Climate variables and soil moisture monitoring. Long-term meteorological data from eld stations close to the study sites were downloaded from the database of the Chinese National Meteorological Information Center/China Meteorological Administration (NMIC/CMA). For this study, ve climate variables (MAP, MAT, MARH, annual days without rainfall, and annual days with small rainfall) were selected.
Soil moisture was measured using both a neutron probe and gravimetric sampling in eight plots located at four stations (DH, HT, BN, and AL). All results were converted into volumetric water content (%) of the respective soil layers aer being combined with soil bulk density data. Moisture located in the top 50cm of the soil was selected as the soil water factor.
For comparison among dierent sites, all of the compiled data were standardized using Equation1:
=
x x x
ij ij ij
where, xij is the standardized data that corresponds to the original data xij, dimensionless; i represents the station; j represents the census year; xij represents the means of the measurement values in the respective durations for the station i; and ij represents the SD of measurement values in the respective durations for station i.
NPP data processing. An increasingly important role in identifying changes in terrestrial NPP is being played by satellite sensors. The annual NPP dataset used in this study was generated from the GloPEM (1981 2000) and MOD17A3 (20002012) data. The GloPEM dataset is derived from Advanced Very High Resolution Radiometer (AVHRR) images at an 8km resolution, as obtained from the AVHRR Path nder Project. As a part of NASAs Earth observatory System (EOS) program, MOD17 MODIS has produced the rst satellite derived dataset to recorded vegetation productivity at a global scale. Here we used the MOD17A3 annual NPP data at a 1km resolution and the annual NPP data from GloPEM. Due to both NPP data sources containing the same year (2000), a relative correction method was used to remove the dierences. First, to keep the same resolution and better accuracy, the MOD17A3 data (1km resolution) were resampled into an 8km resolution using the bilinear interpolation method. A number of group samples were then extracted in the same position nearby the study area from the two processed NPP data sets (2000) which were used to analyze correlations. The regression equation (y = 0.278x + 590.8, R2 = 0.56) was used to adjust the MOD17A3 data. Then, as the nal step, the NPP values (20012012) were extracted from the adjusted MOD17A3 data, and the NPP values (19812000) were directly extracted from the GloPEM NPP data for each of the ve stations. All the data processing was performed using ARCGIS (version 10.1, ESRI) soware.
Statistical analysis. A simple linear regression model was used to regress the trends of standardized data representing NPP and climate variables for the ve sites. Person correlation analysis was performed to identify
ij
(1)
SCIENTIFIC REPORTS
7
www.nature.com/scientificreports/
the relationship between the standardized data representing annual NPP and climatic variables; the contributions of the standardized data representing climatic variables on NPP were calculated with stepwise multi-regression analysis. Statistical analysis and data processing were conducted using SPSS (version 19.0, SPSS Inc.) soware.
References
1. Horst, C. P. & Munguia, P. Measuring ecosystem function: consequences arising from variation in biomass-productivity relationships. Community Ecol. 9, 3944 (2008).
2. Le Qur, C. et al. Trends in the sources and sinks of carbon dioxide, Nature Geoscience 2, 831836 (2009).3. Harte, J. et al. Global warming and soil microclimate: Results from a meadow-warming experiment. Ecol. Appl. 5, 132150 (1995).4. Piao, S. L. et al. Footprint of temperature changes in the temperate and boreal forest carbon balance. Geophys. Res. Lett. 36, L07404 (2009).
5. Nemani, R. R. et al. Climate-Driven increases in Global Terrestrial Net Primary Production from 1982 to 1999. Science 300, 15601563 (2003).
6. Xia, J. Y. et al. Terrestrial carbon cycle aected by non-uniform climate warming. Nature Geoscience 7, 173180 (2014).7. Zhao, M. & Running, S. W. Drought-induced reduction in global terrestrial net primary production from 2000 through 2009. Science 329, 940943 (2010).
8. Piao, S. L. et al. The carbon budget of terrestrial ecosystems in East Asia over the last two decades. Biogeosciences 9, 35713586 (2012).
9. Ciais, P. et al. Europe-wide reduction in primary productivity caused by the heat and drought in 2003. Nature 437, 529533 (2005).10. Wang, X. H. et al. Spring temperature change and its implication in the change of vegetation growth in North America from 1982 to
2006. PNAS 108, 12401245 (2011).
11. NI, J. Estimating net primary productivity of grasslands from eld biomass measurements in temperate northern China. Plant Ecol. 174, 217234 (2004).
12. Piao, S. L. et al. Net carbon dioxide losses of northern ecosystems in response to autumn warming. Nature 451, 49U43 (2008).
13. Nepstad, D. C. et al. The eects of partial through fall exclusion on canopy processes, aboveground production, and biogeochemistry of an Amazon forest. J. Geophys. Res. 107, D8085 (2002).
14. Brando, P. M. et al. Seasonal and interannual variability of climate and vegetation indices across the Amazon. PNAS 107, 1468514690 (2010).
15. Lee, J. E. et al. Forest productivity and water stress in Amazonia: observations from GOSAT chlorophyll uorescence. Proc. R. Soc. B 280, 20130171 (2013).
16. Zhou, G. Y. et al. Substantial reorganization of Chinas tropical and subtropical forests: based on the permanent plots. Glob. Change Biol. 20, 240250 (2014).
17. Zhou, G. Y. et al. Quantifying the hydrological responses to climate change in an intact forested small watershed in Southern China. Glob. Change Biol. 17, 37363746 (2011).
18. Clark, D. B., Clark, D. A. & Oberbauer, S. F. Annual wood production in a tropical rain forest in NE Costa Rica linked to climatic variation but not to increasing CO2. Glob. Change Biol. 16, 747759 (2010).
19. Breshears, D. D. et al. Tree die-o in response to global change-type drought: mortality insights from a decade of plant water potential measurements. Front. Ecol. Environ. 7, 185189 (2009).
20. Allen, C. D. et al. A global overview of drought and heat-induced tree mortality reveals emerging climate change risks for forests. For. Ecol. Manage. 259, 660684 (2010).
21. Ma, Z. H. et al. Regional drought-induced reduction in the biomass carbon sinks of Canadas boreal forests. PNAS 109, 24232427 (2012).
22. Zhou, G. Y. et al. A climate change-induced threat to the ecological resilience of a subtropical monsoon evergreen broad-leaved forest in Southern China. Glob. Change Biol. 19, 11971210 (2013).
23. Nepstad, D. C. et al. Mortality of large trees and lianas following experimental drought in an Amazon forest. Ecology 88, 22592269 (2007).
24. Prince, S. D. & Goward, S. N. Global primary production: A remote sensing approach. J. Biogeogr. 22, 815835 (1995).25. Zhao, M. et al. Improvements of the MODIS terrestrial gross and net primary production global data set. Remote Sens. Environ. 95, 164176 (2005).
26. Gong, D. Y. & Shi, P. J. Inter-annual Changes in Eurasian Continent NDVI and Its Sensitivity to the Large-scale Climate Variations in the Last 20 Years. Acta Botanical Sinica 46, 186193 (2004).
27. Churkina, G. & Running, S. W. Contrasting climatic controls on the estimated productivity of global terrestrial biomes. Ecosystems 1, 206215 (1998).
28. Dai, A. G., Trenberth, K. E. & Qian, T. A global dataset of Palmer Drought severity index for 1870-2002: relationship with soil moisture and eects of surface warming. J. Hydrometeorol. 5, 11171130 (2004).
29. Chen, J. M. et al. Boreal ecosystems sequestered more carbon in warmer years. Geophys. Res. Lett. 33, L10803 (2006).30. The editorial board of vegetation of China. Vegetation of China 1 (Science Press, Beijing, China, 1980).
Acknowledgements
This work was supported by the Strategic Priority Research Program of the Chinese Academy of Sciences (XDA05050205) and the PhD research startup foundation of Henan normal university (5101209171104). We are grateful to all researchers and students that were involved in the related eld surveys. PIs of the ve forest stations are: Deqiang Zhang (DH), Min Cao (BN), Xihua Wang (TT), Silong Wang (HT) and Yiping Zhang (AL).
Author Contributions
W.W. analyzed the data and wrote the manuscript. G.Z. and J.Y. designed the study, proposed the scientic hypothesis. J.W. and X.L. compiled the data.
Additional Information
Competing nancial interests: The authors declare no competing nancial interests.
How to cite this article: Wang, W. et al. Decadal drought deaccelerated the increasing trend of annual net primary production in tropical or subtropical forests in southern China. Sci. Rep. 6, 28640; doi: 10.1038/ srep28640 (2016).
SCIENTIFIC REPORTS
8
www.nature.com/scientificreports/
This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the articles Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/
SCIENTIFIC REPORTS
9
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
Copyright Nature Publishing Group Jun 2016
Abstract
Previous investigations have identified that the effects of climate change on net primary production (NPP) of global forests have varied both spatially and temporally, and that warming has increased the NPP for many forests. However, other factors, such as available soil water for plant growth, could limit these incremental responses to warming. In our investigation we have quantified the responses of NPP of tropical or subtropical forests in southern China to warming and drought stress over the past three decades (1981 to 2012) using data from five forest research stations and satellite measurements. NPP, mean annual temperature (MAT) and annual days without rainfall showed an increase of 0.076 g C m-2 a-2 (standardized), 0.057 °C a-1 (standardized) and 0.067 d a-1 (standardized) during the study period, respectively. However, incremental NPP was deaccelerated at a rate of approximately 20.8% per decade. This deacceleration was primarily caused by a decrease in available soil water which resulted from warming (mainly occurring in winter and autumn) and the changes in rainfall pattern. The result indicates that intensifying drought stress would limit future increases of forest NPP in southern China.
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer